Cardiac Output: Determinants, Regulation, and Measurement
An acceptable arterial pressure can coexist with low cardiac output. A rapidly beating heart can eject little blood with each beat. A ventricle with a high ejection fraction can still provide inadequate forward flow.
Cardiac output depends on both the heart’s ability to eject blood and the circulation’s ability to return it. Measuring that flow helps identify why perfusion is inadequate and whether treatment improves it.
6.1 Cardiac Output, Stroke Volume, and Cardiac Index
Cardiac output is the volume of blood ejected by a ventricle per unit time. It is usually expressed in liters per minute. Stroke volume is the volume ejected with each beat, usually expressed in milliliters.
Consider an illustrative heart rate of 90 beats/min and stroke volume of 70 mL/beat. Multiply the number of beats by the volume ejected with each beat:
90 × 70 = 6,300 mL/min = 6.3 L/min
If stroke volume falls to 35 mL/beat at the same rate:
90 × 35 = 3,150 mL/min = 3.15 L/min
The heart beats equally often in both situations, but output is halved because each beat ejects half as much blood.
Cardiac index adjusts output for body surface area:
Cardiac index expresses flow per square meter of body surface area, in L/min/m². For an illustrative output of 4 L/min and body surface area of 2 m²:
4 ÷ 2 = 2 L/min/m²
This adjustment helps compare flow between people of different sizes. Estimated body surface area does not, however, measure the patient’s actual metabolic demand.
The ventricles operate in series. Blood ejected by the right ventricle passes through the lungs before reaching the left ventricle. If right ventricular output temporarily exceeds left ventricular output, blood accumulates in the pulmonary circulation. If the opposite occurs, pulmonary blood volume falls. Without a significant shunt, average effective outputs must therefore become similar at steady state. Establish which part of this pathway a flow measurement represents.
Regurgitation separates total ventricular ejection from useful forward flow. During mitral regurgitation, for example, part of left ventricular ejection returns to the left atrium instead of entering the aorta. A measurement of total ventricular stroke volume can therefore exceed forward aortic stroke volume. Intracardiac shunts can also make pulmonary and systemic flows differ. Interpret the measurement according to the flow it actually samples.
6.2 Heart Rate: Supporting Flow at a Physiological Cost
Heart rate is regulated by pacemaker activity, autonomic signals, circulating hormones, and the conduction system. During acute stress, sympathetic activity can increase both rate and contractile activity.
If stroke volume stays constant, a faster rate increases output because the heart ejects blood more often. Stroke volume may not stay constant, however. A shorter cardiac cycle leaves less time for filling, especially as diastole shortens. It also reduces the time available for much of left ventricular coronary perfusion. At the same time, more frequent contraction raises myocardial oxygen demand.
The effect depends on ventricular properties and the cause of tachycardia. A patient with impaired relaxation may be particularly sensitive to loss of filling time. A patient with fever may have an appropriate increase in rate that accompanies increased metabolic demand.
Return to the illustrative output of 6.3 L/min at 90 beats/min and 70 mL/beat. Now let heart rate rise to 150 beats/min while stroke volume falls to 35 mL/beat:
150 × 35 ÷ 1,000 = 5.25 L/min
Output has fallen from 6.3 to 5.25 L/min despite the faster rate. The loss of stroke volume exceeds the gain from more frequent beats. These assumed values illustrate the relationship; they do not define a heart rate at which filling fails in every patient.
Rhythm affects filling as well as rate. Atrial contraction moves blood into the ventricle late in diastole. Losing that contribution can reduce filling, particularly when ventricular relaxation is impaired. Atrial fibrillation also creates unequal intervals between beats. The resulting differences in filling time produce different stroke volumes from one beat to the next. One selected beat may therefore misrepresent average flow.
The clinical interpretation must identify why the rate changed. Pain, hypovolemia, fever, hypoxemia, and arrhythmia require different responses. Treating a compensatory tachycardia without addressing its cause can remove support for output. Conversely, a rhythm disturbance may itself be the cause of inadequate flow.
6.3 Stroke Volume: Filling, Contraction, and Ejection Load
Stroke volume depends on the conditions before contraction, the strength of contraction, and the load against which ejection occurs. These are commonly discussed as preload, contractility, and afterload.
Preload refers to myocardial stretch before contraction. When additional filling stretches a preload-responsive ventricle, the Frank–Starling mechanism allows it to eject more blood. As the ventricle approaches the flatter part of its function curve, further filling produces a smaller increase in stroke volume. Filling pressure can continue to rise. One resting pressure cannot reliably predict that response.
Contractility describes the myocardium’s ability to generate force when loading conditions are specified. Greater contractility can allow the ventricle to eject more blood from the same initial filling against the same load. Ischemia, inflammation, and metabolic disturbances can impair this ability. Sympathetic stimulation and some inotropic drugs can increase it. A change in observed ejection alone does not isolate contractility if preload or afterload changes at the same time.
Afterload is the mechanical load opposing ventricular ejection. The ventricle must generate enough pressure to open the outlet valve and continue ejecting into the arterial system. Resistance affects that load, but arterial stiffness and the timing of pressure waves also matter. Ventricular size and wall geometry influence the muscle force needed to generate pressure. The right ventricle is particularly sensitive to increases in pulmonary vascular load.
These determinants interact. Increasing arterial pressure may improve coronary perfusion while also increasing the load on the left ventricle. Giving fluid may improve filling while worsening ventricular distension or congestion. A treatment cannot be evaluated from its effect on one determinant alone.
Ejection fraction describes the proportion of end-diastolic ventricular volume ejected during systole. It depends on loading conditions as well as contraction. A large fraction of a small starting volume can still be a small stroke volume. Conversely, a smaller fraction of a large starting volume can represent more blood. If regurgitation is present, part of that ejection may also flow backward. Ejection fraction therefore cannot establish useful forward flow on its own.
Ejection fraction therefore adds information about ventricular performance but does not replace cardiac output. A complete assessment must consider chamber size, loading, valve function, and measured or estimated forward flow.
6.4 The Heart and Venous Circulation Determine Output Together
The heart can eject only the blood that reaches it over time. The venous circulation, in turn, depends on the heart to maintain the pressure conditions that permit continuing return.
Guyton and colleagues examined venous return across different right atrial pressures. Their experimental work helped establish a framework in which venous return and cardiac performance are considered together. The operating circulation reflects their interaction rather than the action of an isolated pump. 1
In the simplified venous-return model, blood flows from the systemic venous reservoirs toward the right atrium because of a pressure difference. If right atrial pressure rises while upstream pressure stays unchanged, the difference narrows and return falls. Greater resistance along the venous pathway also reduces flow at the same pressure difference. Chapter 7 develops this model and its limitations.
A rise in right atrial pressure may accompany impaired right ventricular ejection. In that setting, additional filling can raise venous pressure without restoring useful flow. The systemic consequences can include congestion despite an apparently full circulation.
A change in vascular tone affects both return and ejection. Arterial constriction can raise the load against which the ventricle ejects. Venous constriction reduces the volume veins hold at low pressure and can move blood toward the heart. This can increase the pressure supporting venous return. Output rises only if the combined changes in filling, contraction, and ejection load allow it to rise.
Monnet and colleagues studied selected patients with septic shock who were preload responsive and had low diastolic arterial pressure. Increasing norepinephrine increased cardiac preload and reduced preload dependence. This physiological study illustrates why a vasopressor can influence filling as well as arterial pressure. It does not establish that output will increase with every vasopressor adjustment. 2
Mechanical ventilation also changes return and ejection. Higher intrathoracic pressure can oppose systemic venous return. Lung inflation can increase right ventricular afterload, while positive surrounding pressure can reduce left ventricular transmural ejection pressure. These effects may occur together. A change in output after intubation or a PEEP adjustment therefore does not establish a change in contractility.
6.5 Interpreting a Low, Normal, or High Output
An output value must be judged against the patient’s requirements. The same flow can be adequate in a resting patient and inadequate during marked metabolic stress.
A low output can reduce oxygen delivery. Its importance depends on arterial oxygen content, regional flow, and tissue demand. Peripheral perfusion, mental status, lactate, and organ function help establish whether the flow is sufficient.
A high output does not guarantee adequate circulation. In vasoplegia, low vascular resistance can keep arterial pressure low despite a large total flow. Some tissues may still receive too little blood because flow is distributed unevenly. Impaired extraction can also limit oxygen uptake. Severe anemia creates a different reason for high output: each unit of blood carries less oxygen, so maintaining oxygen delivery may require more blood flow.
Arterial pressure cannot specify output on its own. It reflects the interaction of flow with vascular properties and downstream pressure. A pressure increase after a vasopressor may occur with an increase, little change, or a decrease in output.
Follow the determinants of output over time. If output stays constant while heart rate falls, stroke volume must have increased, assuming the measurements are valid. If maintaining the same output requires a rising heart rate, stroke volume is falling. Also track vasoactive support. An unchanged output on a higher drug dose can conceal worsening native cardiovascular function.
Before treating a number as abnormal, confirm the measurement method, the clinical state at the time of measurement, and whether the value is indexed. Then connect the result to evidence of perfusion and the suspected mechanism of circulatory failure.
6.6 Measuring Output With the Fick Principle
The Fick principle derives flow from the uptake of a substance and the concentration difference across a system. For systemic oxygen transport, the relationship is:
VO₂ is oxygen consumption in mL O₂/min. CaO₂ and CvO₂ are arterial and mixed venous oxygen contents in mL O₂/dL. The factor of 10 provides the unit conversion.
Suppose measured oxygen consumption is 250 mL/min and the arterial–mixed-venous content difference is 5 mL/dL. Each liter of blood therefore loses 5 × 10, or 50 mL of oxygen as it passes through the systemic circulation. Divide total uptake by the oxygen removed from each liter:
250 ÷ 50 = 5 L/min
Five liters of blood must pass through each minute to account for the measured uptake. If oxygen consumption is overestimated, calculated output is also overestimated when the other inputs stay unchanged.
Direct Fick measurement uses measured oxygen consumption. Estimated Fick methods substitute a predicted consumption value. Both can produce a report with several decimal places, but one uses an actual uptake measurement and the other assumes it. Fever, sedation, muscle activity, and disease can change consumption enough to make that assumption inaccurate.
Opotowsky and colleagues compared thermodilution with estimated Fick measurements in large clinical catheterization cohorts. The methods frequently disagreed and should not be considered interchangeable. These data concern estimated oxygen consumption and should not be taken as a demonstration that a properly performed direct Fick measurement is intrinsically inferior. 3
The venous sample must represent the return from the circulation being assessed. Central venous blood does not have the same mixture as pulmonary arterial mixed venous blood and cannot reliably replace it. Obtain arterial and venous measurements under comparable physiological conditions. The content difference is the denominator of the Fick calculation. When it is small, a small sampling or analytical error can produce a large proportional error in calculated output.
The strength of the principle is conservation of mass. Its practical reliability depends on the quality of each measurement used to apply it.
6.7 Thermodilution: Measuring the Passage of a Thermal Indicator
Thermodilution introduces a known change in temperature into the bloodstream and measures it downstream. The measured temperature change is plotted against time. The resulting curve shows how the thermal indicator passes the detector.
With pulmonary artery thermodilution, fluid of known volume and temperature is injected through the catheter’s proximal port. A downstream sensor detects the temperature change in the pulmonary artery. For the same thermal indicator and valid mixing, faster blood flow dilutes and carries the indicator past the sensor more quickly. The integrated temperature change is smaller, giving a smaller curve area. Slower flow generally produces a larger area. The device uses this inverse relationship to estimate output.
The result depends on accurate delivery and detection of the indicator. Injectate volume and temperature, injection technique, catheter position, respiratory variation, and changes in baseline blood temperature can affect the measurement. Unexpected curves require inspection rather than automatic acceptance.
Stetz and colleagues examined the reliability of thermodilution in clinical practice. Their work demonstrated the importance of measurement variability when comparing outputs. Repeated determinations and attention to technique help distinguish physiological change from variation in the method. Averaging readings cannot correct a systematic technical problem. 4
Some systems use a thermal filament to generate repeated temperature changes and provide a continuous output estimate. The displayed result combines information gathered over time. It can therefore lag behind the patient’s current flow and smooth out a brief change. Check its response time before using it to interpret a short preload maneuver.
Transpulmonary thermodilution detects the indicator in a systemic artery after it has crossed the right heart, pulmonary circulation, and left heart. It can provide output measurements and other derived variables. These additional variables have their own assumptions and should not all be treated as direct measurements of the same compartment.
Significant shunts, regurgitant lesions, and unusual circulatory pathways can complicate indicator behavior. The direction and magnitude of error depend on the arrangement and method. A suspected problem should prompt review of the curves and comparison with the wider hemodynamic assessment.
6.8 Doppler Echocardiography: Flow From Area and Velocity
Doppler echocardiography estimates stroke volume by combining the area of a flow pathway with the distance blood travels through it during ejection. A common approach uses the left ventricular outflow tract, or LVOT.
Doppler measures blood velocity through the outflow tract. Adding those velocities over the ejection period gives the velocity–time integral, or VTI. On the Doppler trace, this is the area under the systolic velocity curve. Velocity multiplied by time gives distance, so VTI is expressed in centimeters. It describes how far the sampled blood column travels during one ejection.
When area is in cm² and VTI is in cm, their product is in cm³, equivalent to mL. Multiplying by heart rate and dividing by 1,000 gives output in L/min.
The area is commonly estimated from an assumed circular cross-section:
For an illustrative LVOT diameter of 2.0 cm, first calculate area:
π × 2.0² ÷ 4 ≈ 3.14 cm²
If VTI is 20 cm, multiply area by the distance traveled:
3.14 × 20 ≈ 62.8 cm³ = 62.8 mL per beat
At 100 beats/min, output is:
62.8 × 100 ÷ 1,000 ≈ 6.28 L/min
The first calculation estimates the size of the flow pathway. The second estimates the blood volume passing through with each beat. The third converts that stroke volume into flow per minute.
Diameter error is amplified because area depends on diameter squared. A 10% overestimate makes the measured diameter 1.10 times its true value:
1.10² = 1.21
Calculated area is therefore 1.21 times the true area, or 21% too high, before any VTI error is considered. Doppler alignment and sample position also affect the estimate. A beam that is not aligned with flow can underestimate blood velocity.
For short-term trending, a relative VTI change can estimate a relative stroke volume change if the outflow area and measurement conditions remain stable. Cardiac output also depends on heart rate. A VTI change should not be interpreted as an identical output change if the rate changes substantially.
Jozwiak and colleagues studied the precision of repeated echocardiographic measurements in 100 patients. The number of averaged beats and the use of the same or a different operator affected the smallest change that could be distinguished reliably. Their results reinforce the need to compare a proposed response with the uncertainty of the actual measurement method. 5
Echocardiography adds more than a flow estimate. It can identify ventricular dysfunction, obstruction, valve disease, and pericardial pathology. Those findings may explain why output is inadequate and can be more useful than an isolated numerical estimate.
6.9 Arterial Waveform Analysis and Other Continuous Estimates
Arterial waveform methods estimate stroke volume from features of the pressure pulse. They use a model that relates the waveform to blood flow and the properties of the arterial system.
Some systems are calibrated against an independent output measurement. Others estimate vascular properties from the waveform and patient characteristics. These approaches cannot be assumed to have identical performance.
The arterial pressure pulse depends on both ejected blood volume and arterial properties. A stiff arterial system develops a larger pressure change for the same added volume than a more compliant system. Changes in tone also alter resistance and wave reflection. The waveform can therefore change even when stroke volume changes little. An output algorithm must account for those changes to keep estimating flow reliably.
Monnet and colleagues compared calibrated pulse contour and uncalibrated waveform approaches in patients with septic circulatory failure. The devices differed in their ability to track output changes after fluid and norepinephrine. This study supports evaluating the specific method in the relevant clinical setting rather than assuming that all pressure-based devices behave alike. It should not be interpreted as a comparison of every subsequent device generation. 6
Waveform quality remains essential. Damping, resonance, catheter obstruction, and artifact can degrade the input. Arrhythmia can also challenge estimation. Calibration and signal checks should follow the requirements of the device and the current physiological circumstances.
Other technologies estimate output using electrical, optical, or ultrasound signals from different sites. Their degree of invasiveness does not establish their accuracy. The relevant questions concern validation, precision, response time, and performance in the patient’s actual condition.
Continuous estimates are particularly useful when they reliably track a change that matters. A smooth trace and frequent updates are insufficient if the signal fails during the vasopressor adjustment or instability being assessed.
6.10 Choosing a Method and Resolving Disagreement
The choice of method should follow the unresolved clinical problem. A brief assessment of ventricular function may require echocardiography. Persistent complex shock may justify more continuous information or invasive measurements. The additional information should have a realistic chance of changing management.
No method should be treated as error free. When two outputs disagree, first establish whether they were measured at the same time and under comparable conditions. Then examine which variables were measured, estimated, or assumed by each technique.
For echocardiography, review the outflow area, Doppler signal, rhythm, and sampling. For thermodilution, review indicator delivery, curve quality, and repeatability. For estimated Fick, examine the assumed consumption and venous sampling site. For waveform methods, review signal quality, vascular changes, and calibration.
The clinical findings provide another check. A value that conflicts with marked hypoperfusion or an obvious mechanical abnormality deserves reassessment. However, clinical impression alone cannot prove that one numerical method is correct. Independent anatomical and physiological information strengthens the explanation.
The PAC-Man randomized trial compared management with and without a pulmonary artery catheter in general ICU practice. It did not demonstrate an overall hospital mortality benefit from catheter-guided management. Insertion complications also occurred. Because treatment was left to clinicians, the study tested a management approach rather than a single standardized treatment algorithm. It does not resolve every question about selected shock populations. 7
Use an output measurement to answer a specific question about the circulation or treatment response. If additional measurements no longer change the explanation or the next decision, continued invasive monitoring may offer little benefit while retaining its risks. Reassess the need as the patient improves or the clinical problem changes.
6.11 Applying Cardiac Output Assessment at the Bedside
Consider an illustrative patient with persistent shock. Heart rate is 90 beats/min. Arterial pressure is maintained with norepinephrine, but peripheral perfusion remains poor. A catheterization report lists an estimated Fick output of 6.0 L/min. Repeated pulmonary artery thermodilution measurements are close to 3.2 L/min.
The disagreement should not be resolved by selecting the more reassuring value. The estimated Fick calculation requires review of its assumed oxygen consumption and the samples used to determine the content difference. Thermodilution requires review of injection technique, curve quality, and potential interference from abnormal circulatory pathways.
Suppose focused echocardiography provides a satisfactory Doppler signal. LVOT diameter is 2.0 cm, giving an area of approximately 3.14 cm². VTI is 12 cm. Calculate stroke volume:
3.14 × 12 ≈ 37.7 mL per beat
At the patient’s heart rate of 90 beats/min:
37.7 × 90 ÷ 1,000 ≈ 3.4 L/min
This independent estimate is closer to thermodilution than to the estimated Fick value. Agreement supports a lower-flow interpretation, although the Doppler estimate has its own measurement uncertainty.
For a body surface area of 2.0 m², divide the thermodilution output by surface area:
3.2 ÷ 2.0 = 1.6 L/min/m²
Interpret that index alongside hemoglobin, oxygenation, perfusion, and metabolic demand. Norepinephrine may be maintaining arterial pressure through increased vascular resistance while useful flow remains inadequate.
The next task is to identify the cause. Ventricular contraction, filling, right ventricular load, valve function, rhythm, and possible obstruction all require consideration. A low output is not automatically an indication for fluid. It is also not sufficient, on its own, to establish a need for an inotrope.
After a targeted intervention, the assessment should include both flow and its clinical consequences. A rise in output accompanied by improved perfusion is more informative than a rise in the displayed number alone. Worsening congestion, tachyarrhythmia, or another adverse effect may limit the value of the intervention.
References
1. Guyton AC, Lindsey AW, Abernathy B, Richardson T. Venous return at various right atrial pressures and the normal venous return curve. American Journal of Physiology. 1957;189(3):609–615. doi:10.1152/ajplegacy.1957.189.3.609.
2. Monnet X, Jabot J, Maizel J, Richard C, Teboul JL. Norepinephrine increases cardiac preload and reduces preload dependency assessed by passive leg raising in septic shock patients. Critical Care Medicine. 2011;39(4):689–694. doi:10.1097/CCM.0b013e318206d2a3.
3. Opotowsky AR, Hess E, Maron BA, et al. Thermodilution vs estimated Fick cardiac output measurement in clinical practice: an analysis of mortality from the Veterans Affairs Clinical Assessment, Reporting, and Tracking (VA CART) Program and Vanderbilt University. JAMA Cardiology. 2017;2(10):1090–1099. doi:10.1001/jamacardio.2017.2945.
4. Stetz CW, Miller RG, Kelly GE, Raffin TA. Reliability of the thermodilution method in the determination of cardiac output in clinical practice. American Review of Respiratory Disease. 1982;126(6):1001–1004. doi:10.1164/arrd.1982.126.6.1001.
5. Jozwiak M, Mercado P, Teboul JL, et al. What is the lowest change in cardiac output that transthoracic echocardiography can detect? Critical Care. 2019;23:116. doi:10.1186/s13054-019-2413-x.
6. Monnet X, Anguel N, Naudin B, Jabot J, Richard C, Teboul JL. Arterial pressure-based cardiac output in septic patients: different accuracy of pulse contour and uncalibrated pressure waveform devices. Critical Care. 2010;14:R109. doi:10.1186/cc9058.
7. Harvey S, Harrison DA, Singer M, et al. Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-Man): a randomised controlled trial. The Lancet. 2005;366(9484):472–477. doi:10.1016/S0140-6736(05)67061-4.